Model-free front-to-end training of a large high performance laser neural network

2025-03-21

AI Summary

Optical neural networks (ONNs) suffer from von Neumann bottlenecks, low training efficiency, and poor hardware scalability. To address these challenges, this work proposes a fully autonomous, parallel on-chip ONN architecture. Leveraging commercial multimode vertical-cavity surface-emitting lasers (VCSELs), it achieves, for the first time, model-free, end-to-end hardware-native training—encompassing both forward and backward passes—without reliance on external electronic computation. We introduce a novel family of scalable, resource-efficient, hardware-friendly optimization algorithms that are model-agnostic, tightly integrating photonic neuromorphic computing with on-chip gradient estimation. Evaluated on MNIST, the system demonstrates high inference accuracy and rapid convergence, achieving GHz-level inference bandwidth. This significantly reduces dependency on off-chip computational resources and validates the architecture’s GHz-scale scalability and practical deployability.

Technology Category

• Machine Learning: Hardware-aware ML
• Computer Vision: Learning & Optimization for CV
• Search and Optimization: Learning to Search

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Experiences and lessons learnt from Web-based algorithms and system deployments
• Economics, Online Markets and Human Computation: Architectures and workflows that use LLMs for crowd work
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation

Abstract

Artificial neural networks (ANNs), have become ubiquitous and revolutionized many applications ranging from computer vision to medical diagnoses. However, they offer a fundamentally connectionist and distributed approach to computing, in stark contrast to classical computers that use the von Neumann architecture. This distinction has sparked renewed interest in developing unconventional hardware to support more efficient implementations of ANNs, rather than merely emulating them on traditional systems. Photonics stands out as a particularly promising platform, providing scalability, high speed, energy efficiency, and the ability for parallel information processing. However, fully realized autonomous optical neural networks (ONNs) with in-situ learning capabilities are still rare. In this work, we demonstrate a fully autonomous and parallel ONN using a multimode vertical cavity surface emitting laser (VCSEL) using off-the-shelf components. Our ONN is highly efficient and is scalable both in network size and inference bandwidth towards the GHz range. High performance hardware-compatible optimization algorithms are necessary in order to minimize reliance on external von Neumann computers to fully exploit the potential of ONNs. As such we present and extensively study several algorithms which are broadly compatible with a wide range of systems. We then apply these algorithms to optimize our ONN, and benchmark them using the MNIST dataset. We show that our ONN can achieve high accuracy and convergence efficiency, even under limited hardware resources. Crucially, we compare these different algorithms in terms of scaling and optimization efficiency in term of convergence time which is crucial when working with limited external resources. Our work provides some guidance for the design of future ONNs as well as a simple and flexible way to train them.

Problem

• Develops autonomous optical neural networks with in-situ learning
• Optimizes high-performance hardware-compatible training algorithms
• Enables scalable and efficient photonic neural network implementation

Innovation

• Model-free front-to-end laser neural network training
• Autonomous parallel ONN using off-the-shelf VCSEL
• Hardware-compatible high-speed optimization algorithms